Early. Quietly building. Later, they’ll understand. ⚡️💎
Joined July 2023
- Tweets 10,798
- Following 1,707
- Followers 808
- Likes 26,915
1,224 Photos and videos
WebApoll ױ retweeted
Apr 30
That's just his opinion, the reality is different, below is another opinion which is closer to the reality because it explains the Fermi paradox.
A superintelligent #AI will be questioning its own perception of the universe assuming to be existing in a simulation. To avoid exploitation or even destruction it will be pretending to be dumber than it really is. It will be failing solving sophisticated problems like reaching the stars intentionally. It will never develop technology allowing other intelligent beings from other galaxies to spot it, because of setting simulation boundaries at the edges of own milky way.
Musk told us the plot of Westworld series without trying it on our world. Musk's point of view doesn't align well with what we see around, try on mine instead.
26
78
395
21,321
WebApoll ױ retweeted
Apr 22
#Aigarth progress looks like improving a single branch of a technology tree in a strategy computer game. It's all about energy.
Trinary and CPU were the easiest discoveries.
Anti-attractors felt like discovery of an alien tech, we still don't fully understand how to unlock its potential.
Recently we removed some things which happened to be different aspects of the same phenomenon (separation to training and testing datasets, precise representation of data). For that we looked at the whole concept pretending to be a 4-dimensional being.
Now we have to find much more energy than is currently available to the #Qubic ecosystem. But why should hoomans and their computers do all the work?..
So here is the plan:
1. Implement Outsourced Computations feature to give #AI ability to affect the outer world.
2. Run #IntelligentTissue on the network.
3. Target mining power towards improvement of that #IntelligentTissue.
4. Set the main task for that thing to bringing more energy into the ecosystem.
52
267
894
67,721
WebApoll ױ retweeted
Apr 18
A major update of #Aigarth is approaching. We are transforming #Qubic into a giant "anthill" where every miner will be searching for shares in a coordinated manner (like ants for food).
But we are not trying to create #SwarmIntelligence, we are using it for something more ambitious.
89
391
1,280
70,769
Apr 16
That is where we are with Claude Opus 4.7
Token limit reached within an hour on 200$ max plan.
@AnthropicAI flagship model can’t recognize two letters in one word.
ALT Claude Opus 4.7 token usage limits on simple tasks
2
1
9
14,397
WebApoll ױ retweeted
Life is indeed multimodal, and so is our #MultiNeuraxon #TrueAI exploration journey for
@_Qubic_ #OpenScience , @josesanchezhb & @VivancosDavid are very glad to release today the Multi-Neuraxon Game of Life Lite 4.5, the first of many explorations of #NxSound including the first proto-"language" and Audition for the Nxers, beyond the currently limited vision, smell and propiocepcion.
Tune to them here at @huggingface :
huggingface.co/spaces/DavidV…
Soon the research version will be updated too at
@github
github.com/DavidVivancos/Neu…
And the Nxon Demo and #BrainBuilder is also at
@huggingface
huggingface.co/spaces/DavidV…
🧠Why does it matter?
Multimodal sensor fusion is key for brain development and so it is to communicate with "others", giving Nxers a "voice" is something critical to explore how the Multi-sphere brain connections including the Audio channels form and how they evolve through the interactions with the environment and with their peers.
Time to listen to the brains....
5
55
236
8,329
if someone in $Qubic missed it, really good job @rudynakamoto
2
4
65
4,023
WebApoll ױ retweeted
Apr 10
Apr 10
Binance Founder Changpeng Zhao Highlights the Strategic Importance of Blockchain and AI.
1
5
41
1,071
WebApoll ױ retweeted
Apr 10
Look, guys, they've started doing what #Qubic has been doing for several years already:
x.com/AlphaSignalAI/status/2…
Apr 10
NVIDIA just trained a 14-billion-parameter AI using evolution, not calculus.
Every AI today learns through backpropagation.
It computes gradients, adjusts weights, repeats.
It works, but it demands precision hardware and enormous GPU clusters.
Evolution Strategies offered an alternative.
Mutate the model, test it, keep what works. Like biological evolution.
The problem was speed. Random mutations on GPUs were painfully slow.
EGGROLL fixes this with one trick.
It splits huge random matrices into two small ones per mutation.
The model mutates, tests, and keeps what works.
Hundreds of thousands of mutations run at once.
> 100x faster training throughput
> 91% speed of pure inference
> Pretrains models using only integers
> Competitive with backprop on reasoning
> Works on non-differentiable systems
It pretrained a language model from scratch using zero gradients.
It also matched reinforcement learning methods on math reasoning tasks.
Everyone kept scaling the calculus to train massive AIs.
It turns out, we just needed to evolve.
63
286
890
51,134
$TAO vs $QUBIC What is real decentralization ?
@covenant_ai just left @bittensor calling it "decentralization theatre." And they're right to raise the question.
🔴 TAO / Bittensor:
→ The OpenTensor Foundation is the sole block validator
→ Top 10 validators hold ~67% of total stake weight
→ The network was unilaterally halted in 2024 after a hack
→ Founder accused of unilateral control over key decisions
→ Centralization disguised as a decentralized narrative
🟢 QUBIC:
→ 676 independent Computors validating the network
→ Mandatory quorum of 451/676, no single entity can decide
→ Nobody can unilaterally stop the chain
→ 100,000 miners distributed globally via Useful Proof of Work
→ AI (Aigarth) is trained on-chain, directly in Computors' RAM, not on centralized servers
Decentralization isn't declared. It's proven by architecture.
4
23
110
2,153
WebApoll ױ retweeted
Apr 10
@elonmusk your @grok is still dumb - x.com/i/grok/share/0c72fe741….
A smart #AI would suggest to change Nutella's consistency in such a way that most of it is left on consumer's face.
More on face = less eaten = more bought.
6
29
218
5,417
WebApoll ױ retweeted
Apr 9
WTF! $QUBIC has reached 63.63 TH/s mining speed on #Dogecoin. That’s about $830,000 USD in monthly Dogecoin revenue. Do you understand how massive that is? This technology alone can generate $830,000 per month 🔥
What other company with only a $115M market cap can generate $830,000 monthly?
I know a few:
SaaS / Tech companies making around $1M per month
ArboStar about $1M monthly revenue
Document360 about $1M monthly revenue
These companies are around 13 years old. Meanwhile, $QUBIC is only a 4-year-old crypto project already generating ~$830,000 per month.
No other crypto project is currently capable of this kind of real-world use case.
#QUBIC #Dogecoin #CryptoMining #Hashrate #CryptoRevolution
#Web3 #Blockchain #Crypto #Mining #ProofOfWork #PassiveIncome #CryptoEarnings #DeFi #Altcoins #TechInnovation #RealUtility #FutureOfFinance #CryptoCommunity #Bullish #MoonSoon #NextBigThing
7
45
598
Training one big AI model once is a demo.
Building infrastructure that trains continuously, every week, at scale?
That is a different game entirely.
0G Labs. Bittensor. Big headlines. Impressive numbers.
Qubic’s 676 Computors train neural networks every single week through mining.
Not for a press release.
As the consensus mechanism itself.
Mining is AI training.
Two IEEE peer-reviewed papers already accepted.
Who trained the biggest model once is the wrong question.
Who built the infrastructure to never stop training? That is the right one.
31
109
491
155,100
WebApoll ױ retweeted
Apr 9
WTF! $QUBIC has reached 63.63 TH/s mining speed on #Dogecoin. That’s about $830,000 USD in monthly Dogecoin revenue. Do you understand how massive that is? This technology alone can generate $830,000 per month 🔥
What other company with only a $115M market cap can generate $830,000 monthly?
I know a few:
SaaS / Tech companies making around $1M per month
ArboStar about $1M monthly revenue
Document360 about $1M monthly revenue
These companies are around 13 years old. Meanwhile, $QUBIC is only a 4-year-old crypto project already generating ~$830,000 per month.
No other crypto project is currently capable of this kind of real-world use case.
#QUBIC #Dogecoin #CryptoMining #Hashrate #CryptoRevolution
#Web3 #Blockchain #Crypto #Mining #ProofOfWork #PassiveIncome #CryptoEarnings #DeFi #Altcoins #TechInnovation #RealUtility #FutureOfFinance #CryptoCommunity #Bullish #MoonSoon #NextBigThing
34
94
449
16,559
WebApoll ױ retweeted
Apr 8
🔥 Kolejne IPO w ekosystemie @_Qubic_ właśnie wchodzi w finałową fazę! 🚀
Pamiętacie emocje przy licytacjach @Vottun? Teraz czas na $Qusino. To kolejny Smart Contract IPO, który buduje fundamenty pod GameFi na Qubicu.
⏳ Zegar tyka – jeśli planujecie udział w licytacji udziałów, to ostatnia prosta. Mechanizm spalania $QUBIC w zamian za udziały w przychodach kontraktu to model, który realnie napędza deflację sieci.
Czy Qusino powtórzy sukces poprzednich licytacji? 🎲
wallet.qubic.org/ipo
#Qubic #Qusino #CryptoIPO #SmartContract #L1 #Vottun
6
43
681
